Monitoring of mining-induced fractures in overburden based on BPNN full-space resistivity inversion
Mining-induced fractures in overburden(MOFs)are good channels for water and gas migra-tion,and this study aims to perform full-space resistivity inversion on the evolution of MOFs monitored by the borehole resistivity method.First,representative"two-zone"geoelectric models of MOFs were constructed regarding different monitoring stages of the borehole resistivity method in the mining face.The apparent resistivity data calculated from different models by the finite element method were employed as the training set and test set of the back propagation neural network(BPNN)for full-space resistivity inversion.On this basis,full-space resistivity inversion tests were carried out on the trained network with the test set,and a test model was established to comparatively analyze the results of BPNN full-space resistivity inversion,the results of resistivity inversion by the least square method,and the fit-ting results of resistivity data by the test model.Finally,the effect of the BPNN full-space resistivity in-version method was tested by the field measurement results of the maximum heights of mining-induced fractures in the overburden of the 9-204 working face of a mine in Shanxi.The research results suggest that the BPNN full-space resistivity inversion method can accurately identify the location and range of MOFs and distinguish MOFs above and below the borehole.This study realized full-space resistivity in-version on the evolution of MOFs monitored by the borehole resistivity method.The research findings can provide technical support for the accurate monitoring of MOFs by the borehole resistivity method.
mining-induced fractures in overburdenborehole resistivity methodmonitoringBPNNfull-space resistivity inversion
万方数据
维普
